- Top of page
- Materials and Methods
The Notch proteins constitute a family of transmembrane receptors that play a pivotal role in cellular differentiation, proliferation and apoptosis. Although it has been recognized that excess Notch signaling is potentially tumorigenic, little is known about precise mechanisms through which dysregulated Notch signaling induces neoplastic transformation. Here we demonstrate that Notch signaling has a transcriptional cross-talk with transforming growth factor-β (TGF-β) signaling, which is well characterized by its antiproliferative effects. TGF-β-mediated transcriptional responses are suppressed by constitutively active Notch1, and this inhibitory effect is canceled by introduction of transcriptional coactivator p300. We further show that this blockade of TGF-β signaling is executed by the sequestration of p300 from Smad3. Moreover, in a human cervical carcinoma cell line, CaSki, in which Notch1 is spontaneously activated, suppression of Notch1 expression with small interfering RNA significantly restores the responsiveness to TGF-β. Taken together, we propose that Notch oncoproteins promote cell growth and cancer development partly by suppressing the growth inhibitory effects of TGF-β through sequestrating p300 from Smad3. (Cancer Sci 2005; 96: 274 –283)
The Notch pathway constitutes an evolutionarily conserved signaling pathway that mediates critical cell fate decisions, such as differentiation, proliferation and apoptosis.(1,2) In addition to the fact that Notch signaling plays pivotal roles in embryonic development, and post-embryonic growth and differentiation in multiple systems such as the immune system,(3,4) substantial evidence indicates that the constitutively activated forms of Notch family proteins are involved in tumorigenesis:(5–17)Notch1/TAN-1 was originally identified as a recurrent chromosomal translocation, t(7;9)(q34; q34.3), in a human acute T-cell lymphoblastic leukemia,(5) resulting in the expression of an extracellular region-truncated form of Notch1 that is known to be constitutively active. The N-terminal truncated forms of the Notch1 and Notch2 proteins have been implicated in the transformation of rat kidney cells in cooperation with an adenoviral oncoprotein, E1A.(10,15) Activated forms of the Notch1(8) and Notch3 proteins(13) are capable of generating T-cell leukemia when retrovirally introduced into bone marrow cells that are transplanted into irradiated recipient mice. Also, Int-3, which encodes a truncated form of Notch4, has been shown to contribute to the generation of mammary carcinoma in mice.(6,11) Interestingly, recent reports have suggested that Notch1 is upregulated in Ras-transformed cells in which activation of Notch1 signaling is necessary to maintain the neoplastic phenotype.(18) Notch activation that causes human neoplasms has been shown to result not only from the truncation, based on the genetic aberration, but also Notch ligand stimulation,(18,19) suggesting that Notch activation without its own genetic abnormalities could be frequently involved in tumorigenesis.(20,21)
Despite rapidly accumulating information about the Notch signaling system, little is known about the mechanism through which excess Notch signaling triggers cellular transformation. One of the clues to this issue is the fact that Notch serves as an adaptor for molecules involved in transcriptional machinery, among which we focus on p300,(22) one of the most common transcriptional coactivator proteins.
The p300 protein interacts with molecules functioning in multiple signaling pathways. Transforming growth factor-β (TGF-β) also uses p300 through activated Smad3.(23,24) TGF-β inhibits proliferation of a wide range of cells including epithelial, endothelial and hematopoietic cells. It plays an important role in controlling tumor development, and its signaling constitutes one of the tumor-suppressor pathways.(25–27) Smads are a class of proteins that function as intracellular signaling effectors for the TGF-β surperfamily, which includes TGF-β, activins and bone morphogenetic proteins (BMP).(28,29) Smad2 and Smad3 are directly phosphorylated by the type I TGF-β receptor in response to TGF-β, leading to formation of heteromeric complexes with Smad4, and are then translocated into the nucleus where they bind to the TGF-β-responsive regulatory sequences, either directly through the Smad-binding elements or in conjugation with other sequence-specific DNA-binding proteins.(30–32) It is suggested that p300 forms the bridge between the Smad complex and the transcriptional apparatus.
Here we show that constitutively active Notch1, consisting of the intracellular domain alone (ICN1), inhibits the antiproliferative activity of TGF-β via the sequestration of p300 from Smad3. We propose that conferring resistance to TGF-β signaling may, in part, be attributed to a mechanism of Notch-induced neoplastic transformation.
- Top of page
- Materials and Methods
In this study, we have demonstrated a transcriptional cross-talk between the Notch and TGF-β signaling pathways. Because Smad proteins are important tumor suppressors, the ability of active Notch1 (ICN1) to repress TGF-β signaling could be responsible, at least partially, for the transforming activity of Notch. A recent study has reported that ICN1 blocks TGF-β-mediated growth arrest in epithelial cells.(41) In that context, ICN1 deregulates expression of c-Myc and thereby renders epithelial cells resistant to growth-inhibitory signals, suggesting a novel link between Notch and cell cycle control. In the experiments described here, we show another mechanism explaining the antagonism between the Notch and TGF-β signaling systems, that is, repression of TGF-β-mediated signaling through sequestration of coactivator p300 by ICN1, which is apparently independent from the mechanism demonstrated by Rao and Kadesch.(41)
Importantly, some investigators have demonstrated that Notch and Smad signaling show functional synergism.(42–44) More complexly, transcriptional activation of the hairy/enhancer of split (HES)-related gene Hey1 is both a direct target of Smad3 and an indirect target through Smad3-dependent transcriptional activation of Notch signaling component genes.(45) Demonstration of direct and TGF-β-dependent interactions between Smad3 and ICN1,(44) and Smad1/5 and ICN1,(42,43) indeed serves as bona-fide evidence of the cross-talk between these two signaling systems. It appears that various molecular interactions could exist between these two signaling systems, most likely in a cell context-dependent manner. Indeed, there is a report showing that both synergy and antagonism could occur between the Notch and Smad signaling systems.(43)
Many transcription factors, including ICN1 and Smads, use the coactivator p300 to activate transcription.(22–24) The p300 protein is generally present at limiting concentrations within the cell nucleus, and functional antagonism between transcription factors occurs as a consequence of direct competition for binding to p300.(46–50) Domains within the p300 protein for interaction with individual transcription factors are highly variable, but both active Notch1 and Smad have been reported to bind to the C-terminal domain, which can potentially be shared. Our results suggest competition between active Notch1 and Smad for limiting quantities of complexes containing p300. Similar competition for p300 has been described for several cellular pathways, including nuclear receptor and AP-1,(46) p53 and E2F,(47) NF-κB and p53,(48) NF-κB and nuclear receptor,(49) and STAT and AP-1.(50)
Regarding Notch-induced transformation, previous studies have indicated that in baby rat kidney cells (RKE) immortalized with E1A, the minimal transforming domain includes ANK and flanking 107 C-terminal amino acids.(15) Consistent with this, our data showed that the EP domain, adjacent to ANK, is required for suppressing Smad activity by ICN1. Recently, the crystal structure revealed that the LDE motif in the EP domain not only governs the stability around this domain but also potentially contributes to direct contacts with p300,(51) supporting our result that the EP mutant (LDE/AAA) fails to sequester p300 from Smad3. Moreover, it is interesting that p300 was isolated as a cellular target of the adenoviral oncoprotein E1A,(52) which is known to block the functions of p300. Therefore, we can speculate that ICN1 may promote sequestration of p300 from Smad3 in cooperation with E1A in RKE cells.
In this study, the biological phenomenon under this transcriptional cross-talk was assessed by both upregulation and downregulation of Notch signaling. For the former, we used strategies of ligand stimulation and overexpression of constitutively active Notch1. For the latter, siRNA-based suppression of Notch1 synthesis in CaSki cells was used successfully to significantly reduce spontaneous generation of the cleaved (i.e. active form of) Notch1. We can speculate that Notch1 is spontaneously activated in CaSki cells either by ligand stimulation from neighboring cells or a cell-autonomous mechanism. In either case, we have demonstrated that spontaneous activation of Notch1 contributes to the growth of CaSki cells, and that blockade of this activation results in a recovered responsiveness to TGF-β. Taken together, we have here demonstrated that active Notch1 may serve as a positive regulator of cell growth by repressing TGF-β-induced growth inhibition. We observed, however, that CaSki cells made no response to TGF-β under treatment with γ-secretase inhibitors, chemical compounds that block Notch cleavage, despite our observation that the amount of active Notch1 was decreased and the transcriptional activation of the Notch reporter gene was suppressed when CaSki cells were treated with γ-secretase inhibitors (data not shown). This observation was apparently puzzling. However, it has since been reported that many transmembrane proteins, in addition to Notch and the amyloid precursor protein that was the substrate identified originally, could be substrates of the γ-secretase.(53) These new lines of evidence made it possible for us to speculate that γ-secretase inhibitors might influence other growth signal pathways and that the specific knockdown of Notch signaling might be achieved using the RNA interference technique, rather than with a γ-secretase inhibitor. It is of future interest to elucidate the mechanisms underlying the failure to restore responsiveness to TGF-β by γ-secretase inhibitors in CaSki cells.
Genetic and molecular studies have implicated several downstream components in the Notch signaling pathway, such as RBP-J and Deltex. As RBP-J is one of the main effectors in Notch signaling,(37,54) it is critical to determine whether the RBP-J-dependent transcription is required for the inhibitory effect of Notch signaling. If that is the case, DN-RBP, a DNA-binding mutant that perturbs Notch activity in a dominant-negative manner, should cancel this suppression. The negative result of the experiment using DN-RBP, however, suggests that RBP-J-dependent transcription of specific target genes is not required for the inhibition of TGF-β signaling.
The p300 protein functions as global transcriptional coactivator and plays important roles in a broad spectrum of biological processes, including cell proliferation and differentiation.(52) A role for p300 in tumor suppression has been proposed, and biallelic mutations of p300 have been identified in certain types of human cancers.(55,56) Furthermore, it was reported recently that reintroduction of wild-type p300 suppresses the growth of p300-deficient carcinoma cells.(57) Insufficiency of cyclic AMP response element binding protein (CBP), a coactivator closely related to p300, also results in both Rubinstein–Taybi Syndrome in humans, a disease characterized by an increased propensity for malignancies, and an increased incidence of leukemias in mice, suggesting that characteristics of tumor suppressors may be common to these general coactivators p300 and CBP.(52)
In summary, we propose that activated Notch represses TGF-β-mediated signaling possibly through sequestration of coactivator p300, which contributes to the mechanisms of Notch-induced neoplastic transformation. Our current results indicate that Notch oncoproteins promote cell proliferation and tumor development partly by repressing the tumor suppressor Smad.